On 23rd July Research England published draft guidance on REF2021 submissions and draft panel criteria and working methods, which contained new information about how they are likely to assess impact. Although the new guidance is out for consultation and so may change before it is eventually finalised in “early 2019”, here are eight things we now know about impact in REF2021:
1. We have some shiny new definitions of impact, significance and reach (though there's nothing new other than the level of clarity they will bring)
2. Continuation case studies are defined as cases where additional impacts have arisen in the current REF window from the same underpinning research as a REF2014 case study. They are welcomed across all panels and there is no need to waste space justifying how they build on previous submissions as long as they demonstrate new impacts in the current #REF2021 period. Panel A have “encouraged” the submission of new rather than continuation case studies, but this inconsistency between panels is likely to be contested in consultation responses. New impacts must be additional to impacts claimed in 2014, and must therefore be in the new REF window. The impacts may be “more of the same” type of impact as the previous case study as long as the additional impacts are in the right period. Underpinning research must have published since 2000. If this means that the case study is based primarily on new research, it would be classified as a new rather than a continuation case study
3. Public engagement is not an impact in and of itself, but a pathway to impact. The ultimate benefits of engagement will need to be demonstrated in terms of their significance and reach
4. Evidence of impact should where possible be independently verifiable. This includes evidence in the form of testimonials, which should: be based on evidence rather than opinion (in panel A); disclose potential conflicts of interest e.g. if they were a research partner (according to panel B); and can only inform grades if they are used in the narrative (in panel D)
5. Stern's idea of allowing public engagement based on body of research by others seems to be rejected, as research by the submitting HEI must have made a material contribution to the impact
6. Indicators of research meeting the 2* threshold (such as peer-reviewed funding) can be included but citation data should not be submitted - it will be provided directly by the REF team to panel A and selected UOAs in panels B and C
7. Annex A (page 95) gives a really useful list of example impacts and the types of evidence that could be used to demonstrate their significance and reach.
8. As expected, case study fields remain pretty much the same and we'll be able to add additional meta-data on researchers and countries involved, funders etc. Bonus is that we get up to 5 pages per case study
Finally, remember this is all still in draft form, for consultation, and may change between now and the publication of final guidance in early 2019.